Skip to main content

Connecting with GitHub Copilot

Complete guide for integrating ProdE with Visual Studio Code and GitHub Copilot through Model Context Protocol (MCP).

Supports streamable-http protocol for MCP integration.

Prerequisites​

Before setting up VS Code integration:

  1. CuriousBox AI Account: Active account with repositories in knowledge layer
  2. VS Code Installation: Latest version of Visual Studio Code
  3. GitHub Copilot: Active GitHub Copilot subscription

Setup Process​

Step 1 : Navigate to MCP Settings​

  • Go to your ProdE account and find the MCP Setup section

alt text

Step 2 : Configuration​

File Location: .vscode/mcp.json (in your project root)

{
  "servers": {
    "prode-codebase-understanding": {
      "type": "http",
      "url": "https://api.prode.ai/code-parsing/v1/mcp/",
      "headers": {
        "Authorization": "Bearer <YOUR_TOKEN>"
      }
    }
  }
}

Configuration Steps​

  1. Copy Configuration: The platform provides a complete configuration with your token automatically included
  2. Open Project in VS Code: Navigate to your project directory
  3. Create .vscode Directory: If it doesn't exist in project root
  4. Create mcp.json File: Add the configuration file in .vscode folder
  5. Insert Configuration: Copy and paste the complete JSON configuration from ProdE platform
  6. Save Configuration: Ensure the configuration is saved properly
  7. Run: Click Run to connect the MCP server
Token Management

The configuration copied from ProdE platform automatically includes your authentication token. If you need to update an existing token (e.g., if your current token stops working), you can copy the new token from the MCP Settings screen and replace it in your existing configuration.

Next: Windsurf Setup | Back to MCP Overview

Last updated on